diff options
Diffstat (limited to 'block')
-rw-r--r-- | block/bio-integrity.c | 4 | ||||
-rw-r--r-- | block/bio.c | 17 | ||||
-rw-r--r-- | block/blk-cgroup.c | 146 | ||||
-rw-r--r-- | block/blk-core.c | 2 | ||||
-rw-r--r-- | block/blk-mq.c | 2 |
5 files changed, 104 insertions, 67 deletions
diff --git a/block/bio-integrity.c b/block/bio-integrity.c index 0436c21db7f2..719b7152aed1 100644 --- a/block/bio-integrity.c +++ b/block/bio-integrity.c @@ -51,7 +51,7 @@ struct bio_integrity_payload *bio_integrity_alloc(struct bio *bio, unsigned long idx = BIO_POOL_NONE; unsigned inline_vecs; - if (!bs) { + if (!bs || !bs->bio_integrity_pool) { bip = kmalloc(sizeof(struct bio_integrity_payload) + sizeof(struct bio_vec) * nr_vecs, gfp_mask); inline_vecs = nr_vecs; @@ -104,7 +104,7 @@ void bio_integrity_free(struct bio *bio) kfree(page_address(bip->bip_vec->bv_page) + bip->bip_vec->bv_offset); - if (bs) { + if (bs && bs->bio_integrity_pool) { if (bip->bip_slab != BIO_POOL_NONE) bvec_free(bs->bvec_integrity_pool, bip->bip_vec, bip->bip_slab); diff --git a/block/bio.c b/block/bio.c index 2a00d349cd68..d6e5ba3399f0 100644 --- a/block/bio.c +++ b/block/bio.c @@ -1831,8 +1831,9 @@ EXPORT_SYMBOL(bio_endio); * Allocates and returns a new bio which represents @sectors from the start of * @bio, and updates @bio to represent the remaining sectors. * - * The newly allocated bio will point to @bio's bi_io_vec; it is the caller's - * responsibility to ensure that @bio is not freed before the split. + * Unless this is a discard request the newly allocated bio will point + * to @bio's bi_io_vec; it is the caller's responsibility to ensure that + * @bio is not freed before the split. */ struct bio *bio_split(struct bio *bio, int sectors, gfp_t gfp, struct bio_set *bs) @@ -1842,7 +1843,15 @@ struct bio *bio_split(struct bio *bio, int sectors, BUG_ON(sectors <= 0); BUG_ON(sectors >= bio_sectors(bio)); - split = bio_clone_fast(bio, gfp, bs); + /* + * Discards need a mutable bio_vec to accommodate the payload + * required by the DSM TRIM and UNMAP commands. + */ + if (bio->bi_rw & REQ_DISCARD) + split = bio_clone_bioset(bio, gfp, bs); + else + split = bio_clone_fast(bio, gfp, bs); + if (!split) return NULL; @@ -2009,6 +2018,7 @@ int bio_associate_blkcg(struct bio *bio, struct cgroup_subsys_state *blkcg_css) bio->bi_css = blkcg_css; return 0; } +EXPORT_SYMBOL_GPL(bio_associate_blkcg); /** * bio_associate_current - associate a bio with %current @@ -2039,6 +2049,7 @@ int bio_associate_current(struct bio *bio) bio->bi_css = task_get_css(current, blkio_cgrp_id); return 0; } +EXPORT_SYMBOL_GPL(bio_associate_current); /** * bio_disassociate_task - undo bio_associate_current() diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c index 9f97da52d006..d6283b3f5db5 100644 --- a/block/blk-cgroup.c +++ b/block/blk-cgroup.c @@ -29,6 +29,14 @@ #define MAX_KEY_LEN 100 +/* + * blkcg_pol_mutex protects blkcg_policy[] and policy [de]activation. + * blkcg_pol_register_mutex nests outside of it and synchronizes entire + * policy [un]register operations including cgroup file additions / + * removals. Putting cgroup file registration outside blkcg_pol_mutex + * allows grabbing it from cgroup callbacks. + */ +static DEFINE_MUTEX(blkcg_pol_register_mutex); static DEFINE_MUTEX(blkcg_pol_mutex); struct blkcg blkcg_root; @@ -38,6 +46,8 @@ struct cgroup_subsys_state * const blkcg_root_css = &blkcg_root.css; static struct blkcg_policy *blkcg_policy[BLKCG_MAX_POLS]; +static LIST_HEAD(all_blkcgs); /* protected by blkcg_pol_mutex */ + static bool blkcg_policy_enabled(struct request_queue *q, const struct blkcg_policy *pol) { @@ -453,20 +463,7 @@ static int blkcg_reset_stats(struct cgroup_subsys_state *css, struct blkcg_gq *blkg; int i; - /* - * XXX: We invoke cgroup_add/rm_cftypes() under blkcg_pol_mutex - * which ends up putting cgroup's internal cgroup_tree_mutex under - * it; however, cgroup_tree_mutex is nested above cgroup file - * active protection and grabbing blkcg_pol_mutex from a cgroup - * file operation creates a possible circular dependency. cgroup - * internal locking is planned to go through further simplification - * and this issue should go away soon. For now, let's trylock - * blkcg_pol_mutex and restart the write on failure. - * - * http://lkml.kernel.org/g/5363C04B.4010400@oracle.com - */ - if (!mutex_trylock(&blkcg_pol_mutex)) - return restart_syscall(); + mutex_lock(&blkcg_pol_mutex); spin_lock_irq(&blkcg->lock); /* @@ -721,8 +718,12 @@ int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol, return -EINVAL; disk = get_gendisk(MKDEV(major, minor), &part); - if (!disk || part) + if (!disk) + return -EINVAL; + if (part) { + put_disk(disk); return -EINVAL; + } rcu_read_lock(); spin_lock_irq(disk->queue->queue_lock); @@ -822,8 +823,17 @@ static void blkcg_css_free(struct cgroup_subsys_state *css) { struct blkcg *blkcg = css_to_blkcg(css); - if (blkcg != &blkcg_root) + mutex_lock(&blkcg_pol_mutex); + list_del(&blkcg->all_blkcgs_node); + mutex_unlock(&blkcg_pol_mutex); + + if (blkcg != &blkcg_root) { + int i; + + for (i = 0; i < BLKCG_MAX_POLS; i++) + kfree(blkcg->pd[i]); kfree(blkcg); + } } static struct cgroup_subsys_state * @@ -833,6 +843,8 @@ blkcg_css_alloc(struct cgroup_subsys_state *parent_css) struct cgroup_subsys_state *ret; int i; + mutex_lock(&blkcg_pol_mutex); + if (!parent_css) { blkcg = &blkcg_root; goto done; @@ -875,14 +887,17 @@ done: #ifdef CONFIG_CGROUP_WRITEBACK INIT_LIST_HEAD(&blkcg->cgwb_list); #endif + list_add_tail(&blkcg->all_blkcgs_node, &all_blkcgs); + + mutex_unlock(&blkcg_pol_mutex); return &blkcg->css; free_pd_blkcg: for (i--; i >= 0; i--) kfree(blkcg->pd[i]); - free_blkcg: kfree(blkcg); + mutex_unlock(&blkcg_pol_mutex); return ret; } @@ -1037,10 +1052,8 @@ int blkcg_activate_policy(struct request_queue *q, const struct blkcg_policy *pol) { LIST_HEAD(pds); - LIST_HEAD(cpds); struct blkcg_gq *blkg; struct blkg_policy_data *pd, *nd; - struct blkcg_policy_data *cpd, *cnd; int cnt = 0, ret; if (blkcg_policy_enabled(q, pol)) @@ -1053,10 +1066,7 @@ int blkcg_activate_policy(struct request_queue *q, cnt++; spin_unlock_irq(q->queue_lock); - /* - * Allocate per-blkg and per-blkcg policy data - * for all existing blkgs. - */ + /* allocate per-blkg policy data for all existing blkgs */ while (cnt--) { pd = kzalloc_node(pol->pd_size, GFP_KERNEL, q->node); if (!pd) { @@ -1064,15 +1074,6 @@ int blkcg_activate_policy(struct request_queue *q, goto out_free; } list_add_tail(&pd->alloc_node, &pds); - - if (!pol->cpd_size) - continue; - cpd = kzalloc_node(pol->cpd_size, GFP_KERNEL, q->node); - if (!cpd) { - ret = -ENOMEM; - goto out_free; - } - list_add_tail(&cpd->alloc_node, &cpds); } /* @@ -1082,32 +1083,17 @@ int blkcg_activate_policy(struct request_queue *q, spin_lock_irq(q->queue_lock); list_for_each_entry(blkg, &q->blkg_list, q_node) { - if (WARN_ON(list_empty(&pds)) || - WARN_ON(pol->cpd_size && list_empty(&cpds))) { + if (WARN_ON(list_empty(&pds))) { /* umm... this shouldn't happen, just abort */ ret = -ENOMEM; goto out_unlock; } - cpd = list_first_entry(&cpds, struct blkcg_policy_data, - alloc_node); - list_del_init(&cpd->alloc_node); pd = list_first_entry(&pds, struct blkg_policy_data, alloc_node); list_del_init(&pd->alloc_node); /* grab blkcg lock too while installing @pd on @blkg */ spin_lock(&blkg->blkcg->lock); - if (!pol->cpd_size) - goto no_cpd; - if (!blkg->blkcg->pd[pol->plid]) { - /* Per-policy per-blkcg data */ - blkg->blkcg->pd[pol->plid] = cpd; - cpd->plid = pol->plid; - pol->cpd_init_fn(blkg->blkcg); - } else { /* must free it as it has already been extracted */ - kfree(cpd); - } -no_cpd: blkg->pd[pol->plid] = pd; pd->blkg = blkg; pd->plid = pol->plid; @@ -1124,8 +1110,6 @@ out_free: blk_queue_bypass_end(q); list_for_each_entry_safe(pd, nd, &pds, alloc_node) kfree(pd); - list_for_each_entry_safe(cpd, cnd, &cpds, alloc_node) - kfree(cpd); return ret; } EXPORT_SYMBOL_GPL(blkcg_activate_policy); @@ -1162,8 +1146,6 @@ void blkcg_deactivate_policy(struct request_queue *q, kfree(blkg->pd[pol->plid]); blkg->pd[pol->plid] = NULL; - kfree(blkg->blkcg->pd[pol->plid]); - blkg->blkcg->pd[pol->plid] = NULL; spin_unlock(&blkg->blkcg->lock); } @@ -1182,11 +1164,13 @@ EXPORT_SYMBOL_GPL(blkcg_deactivate_policy); */ int blkcg_policy_register(struct blkcg_policy *pol) { + struct blkcg *blkcg; int i, ret; if (WARN_ON(pol->pd_size < sizeof(struct blkg_policy_data))) return -EINVAL; + mutex_lock(&blkcg_pol_register_mutex); mutex_lock(&blkcg_pol_mutex); /* find an empty slot */ @@ -1195,19 +1179,49 @@ int blkcg_policy_register(struct blkcg_policy *pol) if (!blkcg_policy[i]) break; if (i >= BLKCG_MAX_POLS) - goto out_unlock; + goto err_unlock; - /* register and update blkgs */ + /* register @pol */ pol->plid = i; - blkcg_policy[i] = pol; + blkcg_policy[pol->plid] = pol; + + /* allocate and install cpd's */ + if (pol->cpd_size) { + list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) { + struct blkcg_policy_data *cpd; + + cpd = kzalloc(pol->cpd_size, GFP_KERNEL); + if (!cpd) { + mutex_unlock(&blkcg_pol_mutex); + goto err_free_cpds; + } + + blkcg->pd[pol->plid] = cpd; + cpd->plid = pol->plid; + pol->cpd_init_fn(blkcg); + } + } + + mutex_unlock(&blkcg_pol_mutex); /* everything is in place, add intf files for the new policy */ if (pol->cftypes) WARN_ON(cgroup_add_legacy_cftypes(&blkio_cgrp_subsys, pol->cftypes)); - ret = 0; -out_unlock: + mutex_unlock(&blkcg_pol_register_mutex); + return 0; + +err_free_cpds: + if (pol->cpd_size) { + list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) { + kfree(blkcg->pd[pol->plid]); + blkcg->pd[pol->plid] = NULL; + } + } + blkcg_policy[pol->plid] = NULL; +err_unlock: mutex_unlock(&blkcg_pol_mutex); + mutex_unlock(&blkcg_pol_register_mutex); return ret; } EXPORT_SYMBOL_GPL(blkcg_policy_register); @@ -1220,7 +1234,9 @@ EXPORT_SYMBOL_GPL(blkcg_policy_register); */ void blkcg_policy_unregister(struct blkcg_policy *pol) { - mutex_lock(&blkcg_pol_mutex); + struct blkcg *blkcg; + + mutex_lock(&blkcg_pol_register_mutex); if (WARN_ON(blkcg_policy[pol->plid] != pol)) goto out_unlock; @@ -1229,9 +1245,19 @@ void blkcg_policy_unregister(struct blkcg_policy *pol) if (pol->cftypes) cgroup_rm_cftypes(pol->cftypes); - /* unregister and update blkgs */ + /* remove cpds and unregister */ + mutex_lock(&blkcg_pol_mutex); + + if (pol->cpd_size) { + list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) { + kfree(blkcg->pd[pol->plid]); + blkcg->pd[pol->plid] = NULL; + } + } blkcg_policy[pol->plid] = NULL; -out_unlock: + mutex_unlock(&blkcg_pol_mutex); +out_unlock: + mutex_unlock(&blkcg_pol_register_mutex); } EXPORT_SYMBOL_GPL(blkcg_policy_unregister); diff --git a/block/blk-core.c b/block/blk-core.c index 82819e68f58b..627ed0c593fb 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -3370,7 +3370,7 @@ EXPORT_SYMBOL(blk_post_runtime_resume); int __init blk_dev_init(void) { BUILD_BUG_ON(__REQ_NR_BITS > 8 * - sizeof(((struct request *)0)->cmd_flags)); + FIELD_SIZEOF(struct request, cmd_flags)); /* used for unplugging and affects IO latency/throughput - HIGHPRI */ kblockd_workqueue = alloc_workqueue("kblockd", diff --git a/block/blk-mq.c b/block/blk-mq.c index f53779692c77..7d842db59699 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -1998,7 +1998,7 @@ struct request_queue *blk_mq_init_allocated_queue(struct blk_mq_tag_set *set, goto err_hctxs; setup_timer(&q->timeout, blk_mq_rq_timer, (unsigned long) q); - blk_queue_rq_timeout(q, set->timeout ? set->timeout : 30000); + blk_queue_rq_timeout(q, set->timeout ? set->timeout : 30 * HZ); q->nr_queues = nr_cpu_ids; q->nr_hw_queues = set->nr_hw_queues; |